Improved tenetobids loading of temporal network files (reshaped correctly).
Event files in tenetobids
get_aux_file new name of get_confound_files.
Force order of columns for TemporalNetwork (fixes #73)
Fixing problems regarding documentation and testing
Changed numpy requirement to 1.20.1
Added community_pairs_norm to networkmeasures.sid
Improved sidecar checking.
Added is_jsonable to utils.
Check_packages added (credit to @jolespin). Modified somewhat as importing to globals within wrapper was not working. This was accepted to main branch then removed.
Session in TenetoBids now included, if present.
Ensuring extensions have leading . when calling pybids for v0.14
Considerable speed increase to teneto.networkmeasures.intercontacttimes.
Fixes the issues when shortest paths were trying to be derived on non existent time points (raised in #60)
Fixing problem with df_to_array when time indices do not start at 0.
Fixing problem with dataframe netshape when starttime !=0 (#61)
coveralls resynced with travis (pip install python-coveralls replaced with pip install coveralls)
make_parcellation extension is no .nii.gz instead of nii.gz for later pybids/templateflow versions
Connectivity matrices in TenetoBIDS, for memory, are saved as 2D representations ((node,node), time) instead of i,j,t format for memory.
Removing wrong cased variables (e.g. netIN and TN from utils)
Neurorequirements removed from requirements
Better what is temporal network tutorial
censor_timepoints documentation is clearer.
Changing raiseIfU variable to rasie_if_undirected in check_input
Updated requirements and fixed outdated version of scipy being installed on clean install using pip.
Fixed header bug in documentation
Fixed that TenetoBIDS bids_filter dicts only contain sub,ses,task and run after calling tnet.run().
Updated remove_confounds documentation to fit with new TenetoBIDS
Allows param dicts in TenetoBIDS functions to be numpyarray by converting input to JSON (fixes #62)
Changing argument names in network measures to 'pertime' and 'overtime' replacing 'time' and 'global' to be clearer.
More detailed documentation to temporal_degree_centrality.
More detailed documentation to temporal_closeness_centrality.
Changing argument values for minimize in shortest_temporal_paths, it is now called: 'temporal_distance' instead of time.
Single instance of version specified in teneto/_version.py.
Output format can be specified in binarize,
TenetoBIDS rewritten. Compatible now with pybids. Easier to use.
censor_timepoints and exclude_runs now in bidsutils as functions (and streamlined) instead of in TenetoBIDS.
Python 3.6 is now required.
Scipy requirements updated.
Tougher coding standards implemented through prospector
Renaming getDistanceFunction to get_distance_function.
Created teneto.neuroimagingtools to contains neuroimaging specific functions.
fmriutils and bidsutils have moved to neuroimagingtools.
Changed distance_func_name to distance_func in volatility.
Instantaneous phase synchrony is now between 0 and 1, fixing potential wrapping issue.
Betweenness centrality now normalizes across time with sigma_jk.
Fix error in #53.
Binarize can accept dataframe in sparse.
Remove_confounds can deal with dataframes.
Fixing inclusion of testdata/dummybids when installing with pip.
Asarray flag in get_network_when returns dense version of network.
Adding CustomBIDS to derivative description json file in TenetoBIDS.
Added deprecation warning to TenetoBIDS
Better docs
Renaming temporalcommunity to communitymeasures for clarity/consistency
teneto.io added
readthedocs.yaml added
Fixing case where, for array input, get_network_when with directed edges dropped duplicates.
Fixed missing bracket in get_network_when (credit to: lcandeago #45)
Added teneto.TenetoWorkflows (name may change to just workflows)
Added promiscuity and persitence to temporalcommunity measures
Updating to fmriprep 1.4.0 updated BIDS naming conventions. fmriprep<1.4.0 no longer compatible.
make_parcellation now uses templateflow and all atlases therein
save_tenetobids_snapshot to export current teneto settings. save_to_pickle (and corresponding load function) have been removed as they are not secure.
NaNs in allegiance, which fixes recruitment and integration. Also added squeeze to input. And bug in integration where for loop would generate error.
Removed self from teneto history.
Some minor TenetoBIDS fixes
Removing unused description_string function from bidsutils
Renaming of dummybids data
Placing custom BIDS formatting in external json config folder.
Removing seaborn as requirement (unused)
Removing function load_parcellation_coords
Removing call to eval and instead getattr in tapered windowed method in genereate_temporalnetwork.
Params for distribution in tapered window method in genereate_temporalnetwork should now be a dictionary not a list.
Schaefer 7 network communities now reference the packaged image (17network mask). Before they references the 7 network mask that was not included.
Added plotedgeweight option to sliceplot for varying edge size.
Minor fixed to TenetoBIDS, teneto.timeseries and teneto.communitydetection
Adding export_history to TenetoBIDS.
Divide resolution of louvain by number of time-points (so gamma is relative to each snapshot)
Default of consensus matrix in community detection set to 0.5.
Changed make_parcellation to accept +OH (for subcortical) and +SUIT (for cerebellar)
Fixed error where tags are set and other folders exist in derivatives
Renaming Schaefer atlas files
Adding Scahefer atlas to data/parcellation_defaults/default.json
Better loading of communities in TenetoBIDS
Renmaed network_communities_ to communitytemplate_
Fixed a bug where SUIT and OH were not loaded in make_parcellation
Better communitytemplate_ tables loaded (OH and SUIT included)
Can specify 7 or 17 communitytemplate size when using schaefer atlas
Allow to apply thresholding within bursty_coeff function
Fixed flexibility inversion bug (values were previously reversed)
Added tctc.
Dataset_description.json added to teneto pipeline in derivatives (TenetoBIDS)
pybids always imported in TenetoBIDS (but validator turned off)
dataset_description exported with export_history (as tenetoinfo.json)
fixing exclusion confound when criteria is based on SD (previously an error was raised)
Added dense compatibility to TemporalNetwork and get_network_when
Added instantaneous phase
Fixing speed problems for temporal network!
Fixing issue in within community estimates for bursty_coef when calc='communities'
Added Schaefer atlas
Added SUIT atlas
Fixed error in temporal_louvain/consensus matrix where labels were getting incorrectly sorted.
Adding temporalcommunity module.
Adding flexibility to temporalcommunity
Removing tenetoBIDS.load_community_data()
Adding communities option to load_data.
Fix to TenetoBIDS.get_pipeline_subdir_alternatives
Removing 'csv' loading from TenetoBIDS (should be tsv)
Fix for nodekwargs{'c'} specification in graphlet slice plot. (Fixes #18)
Fixed a bug in load data of TenetoBIDS when loading networkmeasures from multiple subjects (error was thrown).
Document updates in teneto.timeseries
Fixed a bug in TenetoBIDS confound loading where index_col was first confound.
Making dimord consistency throughout TenetoBIDS (time,node) -> this may change.
Make ijt to be int in process_input from TemporalNetwork (assumes all timestamps are discrete and these are accounted for with metadata (sampling rate and timestart))
Fixed bug in TenetoBIDS's temporal community detection where dataframe was passed instead of TemporalNetwork
TemporalNetwork.N and TemporalNetwork.T forced to be integers.
Improved code quality.
Codacy checks added.
More documentation
Contributers page added in docs
temporal_part_coef renamed to temporal_participation_coeff
added: generatenetwork.rand_poisson
small fixes to tenetoBIDS
matplotlib kwargs added to slice_plot scatter and line parts
added hdf5 possibility to TemporalNetwork
Added utils.io for networkx export of a snapshot.
Renamed teneto.derive package to teneto.timeseries
Renamed teneto.derive.derive to teneto.timeseries.derive_temporalnetwork
Renamed teneto.TenetoBIDS.derive to teneto.TenetoBIDS.derive_temporalnetwork
Rewritten participation coefficient. Fixed potential bug.
Added python-louvain dependency
Added temporal network louvain clustering
teneto.communitydetection imported when importing teneto
allow 2d arrays to be imported into temporalnetwork object
fixed transpose bug in TenetoBIDS makeparcellation when removeconfounds=True
mintor general improvements to TenetoBIDS.removeconfounds
Not yet compatible with HDF5-compatible:, temporal_degree_centrality, volatility, fluctuability, sid, topological overlap, and participation coefficient with negative edges)
Rewritten tenetobids to make more compatible with BIDS derivatives RC1 (note completely compatible yet).
Fixed relative import paths
Subnet argument for certain functions have been removed.
Changing default value of decay argument (temporal_degree_centrality, sid) to 0 instead of None
topological_overlap added to networkmeasures.
documentation improvements.
renamed net variable to tnet in networkmeasures.
added local_variation to networkmeasures.
added TemporalNetwork class.
Added randomseed to binomial.
Fixed layering of edges in circleplot
Fixed error in intercontacttimes where preset network was node,time not node,node
Rewritten intercontacttimes to be calculated through TemporalNetwork class' df-list instead of array.
Rewritten (partially) temporal_degree_centrality to be calculated through TemporalNetwork class' df-list instead of array. (communities still use array)
nLabs arugment changed to nodelabels
tLabs argument changed to timelabels
unit argument changed to timeunit
added cmap to circle_plot
rewritten shortest temporal paths. Now outputs pandas dataframe.
temporal_betweenness_centrality added.
Due to problems with installation of iGraph, temporarily removing communitydetection (commenting out code in TenetoBIDS, communitydetection module is not imported and tests commented.) This is until a more user friendly louvain detection is implemented (and is quick).
Removing main.py and some misc files not used
Added check to TenetoBIDS.load_community_data() for file to exist
Added missing warnings import in networkmeasures
Added more TenetoBIDS functions that use _load_data
Correct caclulation of within-volatility when calc=communities
Changed file_hdr and file_idx in _load_data
Improved matching of confound files and file in TenetoBIDS
Fixed bug for confound reports when windowed method used in TenetoBIDS.derive.
Added analysis tag to directory name of derive vs confound directory. Fixed paths to be absolute image links in html report
BIDS_dir is made to abspath in init of TenetoBIDS
Fixed bug where subjects couldn't be set in init of TenetoBIDS
Allowed ability to reload teneto_bids object when loading from pickle (for development purposes as some information (e.g. history) is lost)
Added communitydetection to TenetoBIDS
Added load_community_data to TenetoBIDS
communitydetection.temporal_louvain_with_consensus can be passed a 2d
Several minor improvements to exclusion_files and exclusion_timepoint in TenetoBIDS
Added TenetoBIDS.load_timelocked_data
TenetoBIDS.make_timelocked_events can now also do raw tvc
Added wait to concurrent.
Fixed bug. derive's loading of FC in JC weighting.
Improved tags in TenetoBIDS when loading data.
Improved timepoint exlcusion and bad_file continuity.
If timepoint exlcusion is spline, now the first and last time-point are not naned.
Added event related displacement (option in networkmeasures.volatility)
Added variable community number per time point compatibility for teneto.networkmeasures.sid and tnet.networkmeasures.temporal_degree_centrality
Other minor selection/load imprvement of tenetoBIDS.
removezeros option added to teneto.networkmeasures.temporal_part_coef
Fixed bug in decay in teneto.networkmeasures.temporal_degree_centrality
Added dummy bids test set to data
Fixed set task and set run bug when raw_data_exists is false.
Added graphlet based thresholding.
Added txt file generation to temporal_exclusion_criteria saying how many time-points are deleted.
rename temporal_exclusion_criteria and file_exclusion_critiera to 'set_timepoint_exclusion' and 'set_file_exclusion'
better importing of modules.
more numpydoc-ing of comments
improved tenetoBIDS.get_selected_files with respect to last_analysis_step
added tenetoBIDS.removeconfounds as standalone function that can still be called through tenetoBIDS.make_parcellation
added a tenetoBIDS._load_files() function to unity io code and increase compatibility through the different funcitons/file formats. However not implemented in all functions yet.
Allow to suppress confound report
Added a temporal participation coefficient to networkmeasures (teneto.networkmeasures.temporal_part_coef)
Allow contact representation in louvain clustering.
Started adding communities argument to replace subnet argument in networkmeasures.
Added DeprecationWarning for removal of subnet argument in Teneto 0.3.5
Added networkmeasures.temporal_part_coef and module_degree_zscore option in teneto.networkmeasures.temporal_degree_centrality.
Added file_exclusion_criteria and temporal_exclusion_criteria to TenetoBIDS
Updated confound_files selection in TenetoBIDS
main added (this may be removed at later date as I don't have an interest in maintaining this).
Added njobs argument to TenetoBIDS (and various other functions therein) for parallel computing.
added nilearn.signal.clean for denoising
removed statsmodels as dependence.
Dockerfile added
Added weight-var and weight-mean options to jackknife correlation.
Added make_static_connectivity in TenetoBIDS
Added from-subject-fc open to jackknife correlation in TenetoBIDS.derive.
Numpydoc adopted in docstrings (teneto.utils, teneto.generatenetwork).
Added bad_subjects and set_bad_subjects() in TenetoBIDS
Added readthedocs and pypi badge
Fixed teneto.networkmeasures.volatility when subnet is given.
Changed argument subnet_id in teneto.networkmeasures.volatility to subnet.
Added possibility to append OH subcortical atlas to make_parcellation.
Added tag option to TenetoBIDS
Changed cfg variable name to params in teneto.utils.
Fixed bug in teneto_degree_centrality where decay and subnet set.
Allow * and + in tag in TenetoBIDS.
clean_community_indexes works with temporal clustering
Added iGraph as requirement
Added teneto.communitydetection.louvain
Added calc option to TenetoBIDS.make_time_locked_events allowing for selecting one of multiple networkmeasures
Corrected bug where timelocked files were not placed in timelocked directory
Changed all forms of time_locked and event-locked to timelocked
Added load_parcellation_data to TenetoBIDS
Added history to TenetoBIDS
Added install_requires to setup.py
Removed older examples.
Added functionality for derivative only BIDS folders (ie if no raw data is present).
Added load/save TenetoBIDS object
Added bids to requirements.txt
Added dependency to seaborn (adde to requirements.txt)
Added confound reports to TenetoBIDS.derive
Added draft version of teneto/bids documentation.
Added Gordon2014_333 parcellation
Renamed parcellations to [firstauthor][publicationyear]_[numberofrois]
Added more documentation to functions
Added networkmeasures.sid
Added forced division by 2 for within (calc: time and subnetwork specified)
Added version number to pipeline name in tenetoBIDS
Added TenetoBIDS.make_time_locked_events and TenetoBIDS.load_network_measure
Added teneto.utils.get_dimord
Fixed the teneto.__version__ that wasn't getting updated (stuck on 0.2.2 for previous versions. So check setup.py in previous versions to be sure what version you were using.)
Updated readthedocs (fixed bugs)
Fixed pybids (instead of bids) in requirements. Removed unnecessary requirements
Reverted back the incorrect fix for #1, added
Added distutils version to requirements.
Fixed bursty_coeff when ICT is empty.
Fix for #1
Added confound_pipeline option to TenetoBIDS.
Added nan-to-median for nans in confounds in removal in TenetoBIDS.make_parcellation
made TenetoBIDS.networkmeasures functional (i.e. saving files, removing an error).
replaced function betai with betainc for scipy 1.0 compatibility.
Corrected vminmax error in plot.graphlet_stack_plot when specifying a string.
Changed the default of vminmax in plot.graphlet_stack_plot to minmax
Fixed documentation error in vminma in plot.graphlet_stack_plot
Correct cmap bug in plot.slice_plot
Added utils.load_parcellation
Folder ./data is also included with teneto, at the moment shen2013_tal parcellation is included there. (This is currently under development). And may change name to specify that this has to do with brain research.
Added decay parameter to degree_centrality
Added additional names to 'multiple temporal derivative' method
corrected bug in bursty_coeff when specifying nodes='all' (previously an error was raised).
Added utils.binarize wiith 'binarize_rdp', 'binarize_percent', and 'binarize_rdp' as options
Added a inputtype item in netinfo dictionary returned by utils.process_input
Modified utils.contact2graphlet to ignore an empty nLabs list.
Added utils.create_traj_ranges
Added the module trajectory with rdp compression.
Added subnet argument to networkmeasures.bursty_coeff so that B is calculated per module.
Changed default of params['report'] in derive.derive to 'no'
Added params['report_path'] to derive.derive.
Corrected default dimension order of derive.derive to 'node,time'
Made report_name parameter to derive.report.gen_report()
Code now follows PEP8
Fixed somwe variable names in stats.shufflegroups
Removed misc.distancefuncitons
All distance funcitons are now through scipy.spatial.distance
Renamed the argument "do" in networkmeasures to "calc".
Renamed the argument "sumOverDim" in networkmeasures.temporal_degree_centrality to "axis".
added function utils.check_distance_funciton_input
Added the module teneto.derive.derive with teneto.derive.postpro_pipeline which inclues: fisher transform, box cox transform and z transform. Can be configured to which ones are used. 5 methods of deriving temporal networks.
Report generation implemented in derive.
Code made more readable in places.
Added withinsubnetwork and betweensubnetwork options for volatility.
Optimized shortest_temporal_path to dramatically increase speed (was looping the same node multiple times in line 70).
Added the function utils.multiple_contacts_get_values
Added possibility to get degree centrality per time point in. teneto.networkmeasures.temporal_degree_centrality
Added the ability to specify vmax and vmin in graphlet_stack_plot
renamed clean_subnetwork_indexes to clean_community_indexes (and now included in initializatin)
added warning message to derive.postpro_boxcox if one edge fails to be normal.
fixed nLabs/tLabs bug in slice_plot
Added subnetwork option to temporal_degree_centrality when do='time''
Changed some types in rand_binomial documentation
Added some more customizeable options to graphlets_stack_plot (border width, grid color, bordercolor etc.)
Added possibility to call "subnetwork" option in volatility
Added function, clean_subnetwork_indexes which network index assignment to range between 0 and max(NrSubnetwork)-1.
Updated some of the examples. Also added a ./examples/previous/vx.x.x where previous versions of examples are listed (examples are not created with every new update).
Added (uncommented) the taxicab_distance in misc.distance_functions option. Won't be added to the documentation of distance functions until I check there is no reason why I commented it. But preliminary testing says it gives the correct answer.
Functions misc.correct_pvalues_for_multiple_testing and misc.corrcoef_matrix are added but not yet implemented in any of the main functions.
Provided clearer documentation in shortest_temporal_path
Add possibility of calculating per time point (or per edge/node - but this takes a tone of time) in volatility
Added possibility of calculating temporal_efficiency per node (either "_from" to "_to")
Improved documentation and added references to rand_binomial.
Fixed bug in graphlet_stack_plot which made white colours have black smudges on them. (Also multiple background colours should theoretically be possible now.)
Added option to remove sharpening filter in graphlet_stack_plot
Added misc and distance_functions (fixing volatility)
Fixed naming of call to temporal_shortest_path in temporal_efficiency,reachability_latency and temporal_closeness_centrality
Added process_input function to cut down on repeating code at beginning of networkmeasures
setup.py has been added for installation (e.g. via pip)
Restructured file structure so that importing teneto has 4 submodules: plot, utils, networkmeasures, generatenetwork.
Functions renamed from camel-case to underscore for python-like code.
More comments added to plotting functions
Docs generated and integration with readthedocs.io
Contact representation field contacts is now numpy array instead of tuple.
graphlet_stack_plot plotting function added.
Examples folder added with several jupyter notebook examples.
Generatenetwork module added. rand_binomial function added which generates a random temporal network.
circle_plot.py created containing circle_plot function (previously in slice_graph.py).
variable vlabs has been changed to nlabs. dlabs has been changed to tlabs.
Field nlabs has been added to contact representation.
slice_plot uses information in contact representation when plotting.
scipy dependency now exists (in graphlet_stack_plot).
removed unnecessary and unused import of networkx
restructured the __init__.py files for better import of teneto.
Measures, misc, plot, utils folders added.
All measures outlined in From static to temporal network theory paper are added. (temporal efficiency, closeness centrality, bursty coefficient, reachability latency, intercontactitmes, shortest temporal path, fluctuability, volatility)
Circle_plot and slice_graph added.